Contact us +91 70212 62656

Centralized Predictive Maintenance of Wind Turbines via IoT Analytics for a Renewable Energy Company

About the customer

Operating as an independent entity and a leader in the renewable energy segment since 1976, the customer came together with one of the world’s largest technology companies’ wind power business in a 2017 merger. The combined operations of the newly formed Europe based company is one of the largest in the world, scaling operations in regions such as India where Renewable Energy is the only sustainable solution to massive development in the power sector. As part of this vision, it acquired several of India’s wind capacity projects, necessitating a full IT overhaul of its pan-India wind farms.

Executive Summary

The customer onboarded 38% of India’s additional wind capacity projects in 2016. To fulfil this scale, it was necessary to centralize operations and maintenance out of the India headquarters for wind farms across the country. However, this needed to be seamlessly implemented in very tight timelines, despite severe on-ground challenges with IT systems and capabilities in remote rural areas of India. SmartCirqls was able to provide a viable solution to simplify this process and ensure centralized real-time monitoring with analytics on-farm performance, turbine health, and external conditions.

Here's what we did
  • Centralized monitoring for pan-India wind farms
  • Customized rules for alarms based on sensor parameters
  • Predicted failure before turbine downtime
  • Reduced maintenance costs by hundreds of thousands of dollars
  • Visualized deviation per parameter from 7-day farm average

The Solution

SmartCirqls unified turbine sensor data from wind farms across India into a single, centralized monitoring console for the HQ Operations Centre.

It was necessary for the success of this project, to consult with local operations teams familiar with the limitations of the landscape and on-ground realities of technology in rural India. The project was kickstarted with an initial POC encompassing 20 turbines x 18 tags x 10-minute intervals, at a single live farm. Eventually, the solution unified 80 sensor parameters per turbine from thousands of turbines across India, in near real-time. We defined deviation classes for each parameter and visualized these deviations against a weekly farm average, successfully predicting the chance of turbine failure more than 20 hours before downtime. Based on the deviation classes, automated alerts were set up to anticipate different levels of severity and escalation.

The manufacturer could proactively send field engineers to investigate issues and repair turbines before a negative incident. This led to massive savings in hardware costs that were earlier caused by multiple-part replacement due to undetected turbine failure. SmartCirqls also ensured monitoring with analytics on farm performance, turbine health, and external conditions. The customer could thus gain visibility into each location, regardless of remoteness or physical inaccessibility.

Challenges

  • Difficulty in centralizing nation-wide wind farm monitoring operations in only a few weeks
  • The extremely remote location of India’s rural hubs, where the customer primarily operated
  • The need for next-gen data analytics to create an “Internet of Things (IoT)” ecosystem
  • Risk of zero connectivity to farm sites, for days on end
  • Dated DB systems on-site, requiring extensive upgrade efforts

Benefits

  • Covered both MSSQL and SCADA data via Kepware for online and offline data
  • Predicted turbine failures more than 20 hours in advance, leading to timely fixes
  • Real-time monitoring for Centralized Farm Operations and Maintenance
  • Simple and stable architecture that could quickly be restored after natural disasters
  • Predictive maintenance activities saved hundreds of thousands of dollars in hardware replacement costs

 

To know how SmartCirqls can help your organization optimize processes through the power of Big Data Analytics, contact us at [email protected].

 

Data Sources
Solution Domains